首页 | 官方网站   微博 | 高级检索  
     

SOC软硬件划分系统中的关键算法
引用本文:何湘竹,陈军波,陈亚光,李栋娜.SOC软硬件划分系统中的关键算法[J].计算机工程与应用,2006,42(14):105-108.
作者姓名:何湘竹  陈军波  陈亚光  李栋娜
作者单位:中南民族大学电子信息学院,武汉,430074
基金项目:中国科学院资助项目;中南民族大学校科研和校改项目
摘    要:设计并实现了SOC软硬件划分系统,搭建了软硬件协同设计的平台并描述了软硬件协同设计的流程。运用多目标遗传算法对目标系统的价格、功耗、速度进行优化,采用了基于Pareto支配的适应值赋值、精英保持和密度计算截断操作的方法进行多目标寻优。针对单任务图描述多CPU系统结构的不足,提出采用多任务图来描述的方法,并提出了MTLS性能评估算法,验证系统软硬件划分的优劣。在对比实验中将NSGA2算法运用到本系统中,结果证明论文的多目标寻优算法获得的非支配解80%比NSGA2的非支配解优。

关 键 词:软硬件划分  多目标优化  SOC  性能评估  调度
文章编号:1002-8331-(2006)14-0105-04
收稿时间:2005-09
修稿时间:2005-09

The Key Algorithms in SOC Hardware-Software Partition System
He Xiangzhu,Chen Junbo,Chen Yaguang,Li Dongna.The Key Algorithms in SOC Hardware-Software Partition System[J].Computer Engineering and Applications,2006,42(14):105-108.
Authors:He Xiangzhu  Chen Junbo  Chen Yaguang  Li Dongna
Affiliation:College of Electronic and Information Engineering,South-Central University for Nationalities,Wuhan 430074
Abstract:A SOC Hardware-Software Partition System is designed and realized in this paper.The platform of SOC partition system is setup and the design flow is described afterward.Multi-objective evolutional algorithm is applied in the system to optimize the parameters of price,power and time.Fitness assignment based on Pareto dominant,elitism and truncation operation are used in the evolutional algorithm.Multi-task graph is proposed since the single task graph is not enough to describe the muhi-processors system,and MTLS algorithm is proposed to evaluate the performance of the partition.In the experiment,NSGA2 is applied in the system and the comparison result obviously shows that the method used in this paper is superior.
Keywords:SOC
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号